dicosar: Testing the equality of Pearson correlation coefficients...

In lhe17/dicosar: Two-sample homogeneity test of Pearson's correlations using saddlepoint approximation and signed root of likelihood ratio statistic

Testing the equality of Pearson correlation coefficients between two groups

Description

These are further details.

Usage

dicosar(x1, x2, delta = FALSE, solver = FALSE)

Arguments

x1	an N1 by P matrix, in which N1 is the number of samples in the first group, and P is the number of variables. If P>2, tests will be run pairwisely.
x2	an N2 by P matrix, in which N2 is the number of samples in the second group.
delta	a logical value. If TRUE, the p-value from the Delta method will also be returned when P=2.
solver	FALSE by default. If solver=TRUE, BBsolve will be used to solve the saddlepoint equations when multiroot fails to converge. Note that BBsolve is much slower than multiroot.

Value

r1: The Pearson correlation coefficient of x1.

r2: The Pearson correlation coefficient of x2.

p: The p-value of the test for the equality of Pearson correlation coefficients using DICOSAR.

code: More information about the convergence, optimization, etc. A negative value indicates a problem. See details.

p_delta: The p-value of the test for the equality of Pearson correlation coefficients using the Delta method.

p_global: A p-value of the global test for the equality of two correlation matrices for each pair of the P variables in the input data. Only returned when P>2.

The value of code

1

No issue found.

-2

Sample size is too small and NA returned.

-3

Too many identical values and NA returned.

-4

At least one of the PCCs is one and NA returned.

-5

The algorithm in DICOSAR does not converge.

2 or 3

The algorithm converges, but the simplified version of DICOSAR (i.e., the first-order approximation of the signed root of the likelihood ratio statistic) is used.

Examples

library(dicosar)
n = 100
x1 = matrix(rnorm(2*n),nrow=n)
x2 = matrix(rnorm(2*n),nrow=n)
dicosar(x1,x2)

lhe17/dicosar documentation built on Sept. 27, 2022, 8:09 a.m.